High Temperature Co-Sintering for Metal Supported-Solid Oxide Fuel Cell Fabrication

Loading...
Thumbnail Image

Date

2019-01-23

Authors

Toor, Sannan Yousaf

Advisor

Croiset, Eric

Journal Title

Journal ISSN

Volume Title

Publisher

University of Waterloo

Abstract

Metal supported-solid oxide fuel cells (MS-SOFC) are third generation solid oxide fuel cells (SOFC). In these cells the primary support is a porous metal upon which the active cell layers (anode, electrolyte, and cathode) are deposited. MS-SOFCs are known for their better mechanical stability, tolerance to redox cycles, and are also cheaper than all-ceramic cells since the conventional support material (one of the active components) is almost entirely replaced by a cheaper metal. Several MS-SOFC fabrication methods have been reported in the literature, including thermal spray deposition on pre-fabricated porous metal support, and tape casting and co-sintering of half-cell layers (metal support and electrolyte). In this thesis a study is done on MS-SOFC’s fabrication using tape casting and high temperature co-sintering in non-oxidizing atmosphere to protect the metal support. The MS-SOFCs studied had SS-430L as metal support, and samarium doped ceria (SDC) or yittrium stabilized zirconia (YSZ) as electrolyte. To use SDC as electrolyte, the co-sintering temperature should be lowered to mitigate the reduction of cerium (IV) to cerium (III) in ceria when exposed to reducing atmosphere at high temperatures. To lower the sintering temperature varying amounts of copper (0 - 5.0 mol%) were used as sintering aid with SDC. X-ray diffraction (XRD), scanning electron microscopy (SEM), energy dispersive x-ray (EDX), and electrochemical impedance spectroscopy (EIS) were used to characterize SDC with and without copper. Upon doping, copper goes in the SDC crystal structure and starts forming an additional copper oxide phase for copper content above 0.1 mol%. The SDC crystal lattice constant decreased from 0.5446 to 0.5427 nm when the copper content was increased from 0 to 5.0 mol%. SEM of sintered samples showed similar grain size (1-2 microns) and shape, and minimum visible surface porosity for samples containing 0.5 mol% and more copper. Even though as copper content is increased, the sintering temperature decreases, the ionic conductivity also decreases. 0.5 mol% copper shows the best compromise, with sintering temperature lowered to 1180℃ compared to 1350℃ for plain SDC, and total conductivity of 0.065 S.cm-1 compared to 0.077 S.cm-1 for plain SDC at 800℃. The amount of additional copper oxide phase is lowest for 0.5 mol% copper co-doped SDC, limiting the effect on SDC’s grain boundary conduction. To use tape casting and co-sintering for MS-SOFC half-cell fabrication, shrinkage analysis of cell components (metal support and electrolyte) were carried out using dilatometry. The shrinkage analysis includes the effect of sintering profile (temperature profile, and sintering atmosphere) on the shrinkage dynamics of cell components. YSZ shows similar shrinkage behavior in inert and reducing atmospheres, however increasing the ramping rate between 1000℃ and 1350℃ shifts the maximum shrinkage rate peak to a higher temperature. Dilatometry analysis showed that reducing atmosphere is more suitable for co-sintering than inert atmosphere, as co-sintering in inert atmosphere leads to formation of silica and alumina in the metal support. In addition to being non-conductive, silica and alumina act as sintering aids that enhance the shrinkage of SS-430L. Hence, upon reaching the co-sintering temperature (1350℃) in inert atmosphere, SS-430L is close to its final shrinkage (⁓ 1% shrinkage left) and is rigid. However, upon reaching the co-sintering temperature in inert atmosphere, YSZ is not as close to its maximum shrinkage (⁓3-5% shrinkage left). Therefore, continued shrinkage of thin YSZ layer on rigid SS-430L surface will lead to cracks and delamination. By increasing the ramping rate during co-sintering in reducing atmosphere the shrinkage of SS-430L can be increased to an extent that it is similar to YSZ shrinkage upon reaching sintering temperature. At a ramping rate of 5.0℃/min, the shrinkage of SS-430L layer and YSZ were 17.1% and 18.1%, respectively as opposed to 8.8% and 17.1% at 2.5℃/min ramping. Since increasing ramping rate shifts the maximum shrinkage rate to a higher temperature for YSZ, upon reaching sintering temperature YSZ is not as close to its maximum shrinkage. For 5℃/min ramping rate upon reaching sintering temperature YSZ has to shrink 2.4% more to get to maximum shrinkage whereas sample with 7.5℃/min ramping rate has to shrink 3.1% to get to maximum shrinkage. This means that the thin YSZ layer is softer upon reaching the sintering temperature at higher ramping rate and will not crack upon continued shrinking. In this study, 7.5℃/min ramping rate gives the best co-sintered half-cells without physical defects. The anode catalyst (NiO/SDC) was infiltrated as a solution in the porous metal support, and the cathode catalyst (LSCF/GDC) was printed on YSZ. Initial cell performance testing showed open circuit voltage of 0.989 V at 700℃ and maximum power density of only 0.5 mW.cm-2 at 700℃. For MS-SOFCs with copper co-doped SDC, 0.5 mol% copper co-doped SDC was used. There is a difference in shrinkage set off temperature for copper co-doped SDC and SS-430L, where shrinkage of copper co-doped SDC starts around 700℃ compared to 1100℃ for SS-430L. This means that during co-sintering, the copper co-doped SDC layer will be close to its maximum shrinkage when the sample reaches 1100℃, which is only when SS-430L starts shrinking. Due to such a difference in shrinkage behaviors, MS-SOFC with copper co-doped SDC were not successfully fabricated. Contrary to the initial hypothesis to lower the sintering temperature, because of shrinkage dynamics of SS-430L, higher co-sintering temperatures may be more suitable.

Description

Keywords

Metal Supported Solid Oxide Fuel Cells, Solid Oxide Fuel Cells, Co-Sintering, Samarium doped Ceria, Yttrium Stabilized Zirconia, Stainless steel 430 L, Shrinkage behavior

LC Keywords

Citation